Fabrication of highly stable glyco-gold nanoparticles and development of a glyco-gold nanoparticle-based oriented immobilized antibody microarray for lectin (GOAL) assay

Article abstract of DOI:10.1002/chem.201405747

The design of high-affinity lectin ligands is critical for enhancing the inherently weak binding affinities of monomeric carbohydrates to their binding proteins. Glyco-gold nanoparticles (glyco-AuNPs) are promising multivalent glycan displays that can confer significantly improved functional affinity of glyco-AuNPs to proteins. Here, AuNPs are functionalized with several different carbohydrates to profile lectin affinities. We demonstrate that AuNPs functionalized with mixed thiolated ligands comprising glycan (70 mol%) and an amphiphilic linker (30 mol%) provide long-term stability in solutions containing high concentrations of salts and proteins, with no evidence of nonspecific protein adsorption. These highly stable glyco-AuNPs enable the detection of model plant lectins such as Concanavalin A, wheat germ agglutinin, and Ricinus communis Agglutinin 120, at subnanomolar and low picomolar levels through UV/Vis spectrophotometry and dynamic light scattering, respectively. Moreover, we develop in situ glyco-AuNPs-based agglutination on an oriented immobilized antibody microarray, which permits highly sensitive lectin sensing with the naked eye. In addition, this microarray is capable of detecting lectins presented individually, in other environmental settings, or in a mixture of samples. These results indicate that glyconanoparticles represent a versatile and highly sensitive method for detecting and probing the binding of glycan to proteins, with significant implications for the construction of a variety of platforms for the development of glyconanoparticle-based biosensors.

Full text of DOI:10.1002/chem.201405747

DOI: 10.1002/chem.201405747
Full Paper
&
Biosensors |Hot Paper|
Fabrication of Highly Stable Glyco-Gold Nanoparticles and
Development of a Glyco-Gold Nanoparticle-Based Oriented
Immobilized Antibody Microarray for Lectin (GOAL) Assay
Li-De Huang, Avijit K. Adak, Ching-Ching Yu, Wei-Chen Hsiao, Hong-Jyune Lin, Mu-Lin Chen,
[a]
and Chun-Cheng Lin*
Abstract: The design of high-affinity lectin ligands is critical
for enhancing the inherently weak binding affinities of mon-
omeric carbohydrates to their binding proteins. Glyco-gold
nanoparticles (glyco-AuNPs) are promising multivalent
glycan displays that can confer significantly improved func-
tional affinity of glyco-AuNPs to proteins. Here, AuNPs are
functionalized with several different carbohydrates to profile
lectin affinities. We demonstrate that AuNPs functionalized
with mixed thiolated ligands comprising glycan (70 mol%)
and an amphiphilic linker (30 mol%) provide long-term sta-
bility in solutions containing high concentrations of salts
and proteins, with no evidence of nonspecific protein ad-
sorption. These highly stable glyco-AuNPs enable the detec-
tion of model plant lectins such as Concanavalin A, wheat
germ agglutinin, and Ricinus communis Agglutinin 120, at
subnanomolar and low picomolar levels through UV/Vis
spectrophotometry and dynamic light scattering, respective-
ly. Moreover, we develop in situ glyco-AuNPs-based aggluti-
nation on an oriented immobilized antibody microarray,
which permits highly sensitive lectin sensing with the naked
eye. In addition, this microarray is capable of detecting lec-
tins presented individually, in other environmental settings,
or in a mixture of samples. These results indicate that glyco-
nanoparticles represent a versatile and highly sensitive
method for detecting and probing the binding of glycan to
proteins, with significant implications for the construction of
a variety of platforms for the development of glyconanopar-
ticle-based biosensors.
[4]
[5]
Introduction
enza virus hemagglutinins, cholera toxins, and dendritic cell
(
DC)-specific intracellular adhesion molecule-3-grabbing non-
[6]
The adhesive power of low-affinity carbohydrate ligands is am-
plified by multivalency, a common feature in protein–ligand
integrins (DC-SIGN) to human immunodeficiency virus (HIV),
are often mediated through binding to terminal glycans on
cell-surface glycoproteins. For the prevention of infection,
blocking of the glycan-binding proteins or lectins may be ben-
eficial for therapeutic intervention of various disease states.
The development of a biosensing platform for the study of
these biological interactions at the molecular level would facili-
tate the elucidation of these biologically significant carbohy-
[
1]
binding and cell–cell adhesion events. Lectins, which are
abundant in nature, are carbohydrate-binding proteins that are
[
2]
neither antibodies nor enzymes. In nature, most of these
glycan-binding proteins exist as oligomers with several carbo-
hydrate-binding sites. Carbohydrate–protein interactions play
an essential role in a variety of pathological and physiological
cellular functions such as cell–cell and cell–matrix interactions,
[7,8]
drate–ligand interactions.
[2]
inflammation, development, fertility, and cancer metastasis.
Accordingly, various multivalent glycoconjugates decorated
on a variety of scaffolds, including peptides, dendrimers, syn-
thetic polymers, and proteins, have been designed to control
The interactions between lectins and host cell-surface carbohy-
drate epitopes also represent the vital first step in the cycle of
host infection by several microbes including bacteria, viruses,
[1]
the valency and spatial display of the ligands. Biomolecule-
functionalized nanomaterials are emerging diagnostic tools
[
2,3]
fungi, parasites, and protein toxins.
For example, proteins
[9]
involved in infectious pathogenesis, such as the human influ-
with unique biological applications in a wide variety of areas.
Of particular interest are the unique optical and electronic
properties of gold nanoparticles (AuNPs), which have been
used as biosensors in nanomedicine, biomedical imaging, and
+
+
[
a] Dr. L.-D. Huang, Dr. A. K. Adak, Dr. C.-C. Yu, W.-C. Hsiao, H.-J. Lin,
Dr. M.-L. Chen, Prof. Dr. C.-C. Lin
Department of Chemistry
[10]
other biodiagnostic applications. Glyco-AuNPs, that is, AuNPs
decorated with a monolayer of thiolated carbohydrate ligands,
can function as excellent nanosized scaffolds similar to the gly-
National Tsing Hua University
Hsinchu-300 (Taiwan)
E-mail: cclin66@mx.nthu.edu.tw
[11]
+
cocalyx-like structure covering the cell surface. In addition,
these nanomaterials offer several advantages compared with
current protein scaffolds, such as greater synthetic control,
[
] These authors contributed equally to this work.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201405747.
Chem. Eur. J. 2015, 21, 3956 – 3967
3956
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
a high degree of solubility and stability in buffered aqueous sol-
Here, we prepared different glyco-AuNPs and demonstrated
that glyco-AuNPs can detect cognate lectins at a low picomo-
lar level. We determined that glyco-AuNPs protected with an
additional monolayer of an amphiphilic linker are stable under
high physiological salt concentrations and are resistant to non-
specific protein adherence in solution and on solid supports
such as a microarray. Moreover, we present glyconanoparticle-
based biosensing with the naked eye of lectins bound to an
oriented immobilized-antibody microarray surface (Figure 1).
[
12,13]
utions, high-density ligands, and high flexibility.
The outer
carbohydrate residues located on the surface of glyco-AuNPs
could function as multiple antennae, thereby greatly enhancing
[14]
their ability to detect glycan-mediated interactions in vitro.
Russell and co-workers reported on the use of mannose
Man)-functionalized AuNPs to examine the binding specificity
(
of a well-known plant lectin, Concanavalin A (Con A), by UV/Vis
[
15]
spectrophotometry. The carbohydrate ligand density on the
surface of AuNPs may affect particle aggregation and the over-
all detection sensitivity. Kataoka and co-workers determined
that PEGylated lactose (Lac)-modified AuNPs with 65% surface
coverage produced optimal RCA120-induced AuNP aggrega-
[
16]
tion. Although Lac-functionalized PEGylated AuNPs exhibit
high dispersion stability in aqueous buffer solutions, these
AuNPs are susceptible to the nonspecific adsorption of pro-
[
16]
teins. In addition, the type of saccharide presented on the
AuNPs and the spacer lengths affect the aggregation of the
[17]
AuNPs and the detection sensitivity for RCA120 significantly.
Notably, these AuNPs enabled the study of carbohydrate-
mediated lectin interactions; however, the investigation was
limited to only one type of carbohydrate and one type of
lectin. To analyze the carbohydrate–protein interactions quanti-
tatively, various glyconanoparticle systems bearing different
carbohydrate epitopes have been described for the investiga-
tion of specific binding interactions through several meth-
[
18]
ods.
These methods include surface plasmon resonance
Figure 1. Illustration of a glyconanoparticle-based oriented antibody micro-
array for lectin sensing (GOAL) assay. The GOAL assay permits direct mea-
surement of carbohydrate–lectin interactions by naked-eye visualization of
the surface-bound protein utilizing glyco-AuNPs on glass slides after silver
enhancement.
[
19]
[20]
(
SPR), quartz crystal microbalance (QCM) techniques, iso-
[
21]
thermal titration calorimetry (ITC), magnetic resonance imag-
[
22]
[23]
ing (MRI), fluorescence measurements, and dynamic light
[
24]
scattering (DLS).
We previously reported that glyco-AuNPs can be used for la-
beling specific proteins on the cell surface through carbohy-
Finally, we demonstrate the feasibility of a functional Ab micro-
array platform for the simultaneous detection of lectins in sam-
ples containing a mixture of analytes. As outlined in Figure 1,
when antibodies are immobilized, surface-functionalized bor-
onic acids (BAs) form, in an oriented manner, cyclic boronate
diesters with the diol groups of the glycans in the non-anti-
genic Fc region. This method essentially prevents the antibod-
ies from binding to the solid surface through their antigen-
binding Fab domains, which may facilitate accessibility to an
incoming antigen. The detection of the bound antigen was vi-
sualized by in situ agglutination using high-affinity glyco-
AuNPs following silver enhancement. Because boronate-medi-
ated immobilization surpassed prior antibody modifications,
this assay may provide significantly improved antigen-binding
ability and sensitivity compared with methods employing
random covalent coupling.
[
13]
drate–receptor interactions. The AuNPs decorated with Man
residues were used to visualize the FimH adhesions on type I
pili of E. coli by transmission electron microscopy (TEM). Glyco-
AuNPs exhibit a strong glycoside cluster effect for presenting
carbohydrate ligands that enhance the binding affinity of the
carbohydrate to lectin and are potent anti-adhesives for the
[
11,18]
prevention of bacterial pathogen invasion.
In this context,
we made quantitative measurements of the binding affinities
between the model lectin Con A and various AuNPs carrying
[
19]
multiple copies of Man residues using SPR. Moreover, we de-
k
termined that globotriose (Gb3 or blood group antigen, P )-
coated AuNPs with different particle sizes containing 60 to
1
970 sugars on the surface are inhibitors of the B subunit of
[
25]
Shiga-like toxin (Slt-1). SPR analysis revealed that the relative
potencies of each sugar ranged from approximately 1300 for
a 4 nm particle with a shorter linker to approximately 228000
for a 20 nm particle with a longer linker. Thus, glyco-AuNP ag-
gregation could be tuned for use as a newer screening Results and Discussion
method for monitoring the bond-forming activities of en-
Synthesis of thiolated glycans 1–4 and 6–11
zymes. For example, we recently demonstrated a high-
throughput AuNP-based colorimetric assay for the analysis of
the specific bond formation activity of a model enzyme, a-2,8-
polysialyltransferase, which catalyzes the biosynthesis of a-2,8-
[27]
[28]
Thiolated lactose derivatives (1
and 2 ) equipped with
a shorter linker or a short alkane-thiolate-terminated lactoside
3 (Figure 2) were synthesized as described previously or in the
Supporting Information. To impart flexibility and solubility to
[26]
linked polysialic acid.
Chem. Eur. J. 2015, 21, 3956 – 3967
www.chemeurj.org
3957
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 2. Structures of thiolated glycans (1–4 and 6–11) and a short tri(ethylene glycol) terminated alkanethiol (12) used for the synthesis of glyco-AuNPs.
the glyco-AuNPs, we incorporated an amphiphilic mixed ali-
phatic/ethylene glycol spacer 13 into the neoglycoconjugates.
The p-nitrophenyl carbamate-protected neoglycoconjugates
could be conjugated easily to the amino-functionalized saccha-
rides through a urea linkage. Compound 4 was synthesized
through a procedure developed in our laboratory, and the
versatility of the urea linkage formation was extended to other
glycan structures (15–19; Scheme 1a). Briefly, catalytic hydro-
genation of the azides (14–19) generated the amines, which,
without further purification, were subjected to reaction with
citrate molecules on the AuNP surface were replaced with thio-
lated carbohydrates through a ligand-exchange reaction and
AuÀS bond formation to give glyconanoparticles (Scheme 1b).
The ligand-exchange reaction enabled not only 100% decora-
tion of AuNPs with a single thiolated glycan but also the as-
sembly of two different thiolated ligands on the AuNPs with
varying glycan ratios. Glyco-AuNPs (1–AuNP to 4–AuNP) anch-
ored with thiolated lactose derivatives (1–4) were prepared to
investigate the effect of spacer length on the stability of the
resulting AuNPs in solutions of high ionic strength (100 mm
NaCl). The glycan surface densities on the glyco-AuNPs (5–
AuNP to 11–AuNP) were varied by using solutions consisting
of a mixture of glycans 4–11 (70 mol%) and an amphiphilic
linker 12 (30 mol%). The glyco-AuNPs (1–AuNP to 11–AuNP)
were purified by repetitive centrifugation and resuspension in
either 10 mm Tris buffer (pH 7.4) or 10 mm phosphate-buffered
saline (PBS; pH 7.4). The glyco-AuNPs were fairly dispersed in
aqueous buffer solutions and displayed a redshifted surface
plasmon resonance (SPR) peak at 525 nm, which was associat-
ed with the attachment of thiolated ligands, indicating no self-
aggregation upon surface modification with saccharide (see
the Supporting Information, FigureS2). This small (5 nm) peak
shift is attributable to the formation of a dense organic layer
by the replacement of citrates on the AuNP surface, which cor-
relates significantly with the dielectric constant of the medium
[
26]
13 (structure shown in Figure 2) to give the corresponding gly-
cans as thioacetates. Finally, deacetylation under either Zem-
plꢁn conditions or aqueous NaOH afforded thiolated glycan
derivatives (4, 6–8, 10, and 11) suitable for anchoring on
AuNPs in good yields. This approach provides an easy-to-
handle and efficient technique for installing thiols at the reduc-
ing ends of diverse glycans (Figure 2). In addition, we prepared
a tri(ethylene glycol)-terminated alkanethiol (12; see the Sup-
porting Information for details of the synthesis) and mixed it
with a thiolated glycan to fabricate water-soluble glyco-AuNPs
with minimal nonspecific binding to proteins. Furthermore, the
AuNPs with a self-assembled monolayer (SAM) of poly(ethylene
glycol) exhibited minimum nonspecific adsorption of pro-
[
29]
teins.
[30]
surrounding the nanoparticles.
Preparation of glyco-AuNPs
All of the glyco-AuNPs were synthesized and characterized fol-
[
25,26]
lowing established procedures.
Briefly, the chemisorbed
Chem. Eur. J. 2015, 21, 3956 – 3967
www.chemeurj.org
3958
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 1. Synthesis of a) thiolated glycans (4, 6–8, 10, and 11), and b) glyco-AuNPs (1–AuNP to 11–AuNP). For detailed structures of the thiolated glycans
used for glyco-AuNP synthesis, see Figure 2.
Effect of spacer length and carbohydrate ligand density of
glyco-AuNPs on their stability in high salt concentrations
and resistance to nonspecific protein adsorption
AuNPs are prone to aggregation in solutions of high ionic
[
31]
strength or under physiological conditions.
Therefore, for
biological use, AuNPs must exhibit long-term stability without
nonspecific aggregation in solutions containing high concen-
trations of proteins and salts. For the determination of the
nonspecific agglomeration of AuNPs, the colloidal dispersion
stability of glyco-AuNPs (1–AuNP to 5–AuNP) was assessed
under physiological conditions (Figure 3 and Figure S3). The
short spacers on the AuNPs provide increased colorimetric de-
[
17]
tection sensitivity but also increase aggregation. Upon expo-
sure of nanoparticles (1–AuNP to 3–AuNP) to a solution with
a high NaCl concentration (12.5–600 mm), immediate aggrega-
tion was observed, indicating that these AuNPs are unsuitable
for lectin-sensing applications. In contrast, 4–AuNP containing
a linker with a lipophilic undecyl unit and a tri(ethylene glycol)
chain exhibited very little change in the surface plasmon ab-
sorption band under the same conditions (Figure 3a, dotted
line), suggesting superior stability to nanoparticles 1–AuNP to
3
–AuNP. The slight change in the intensity of the plasmon
band may be attributed to imperfect SAM formation on the
nanoparticle surface owing to steric hindrance between the
[
32]
Figure 3. The effects of different aglycone lengths and ionic strengths on
Lac head groups. Thus, to reduce the hindrance to the for-
mation of a perfect SAM on the AuNPs, we incorporated an
amphiphilic linker (12) to give 5–AuNP. In the high-salt solu-
tions (NaCl concentration from 12.5 to 600 mm), 5–AuNP re-
mained stable, and no change in the absorption intensity was
observed (Figure 3a, solid line). These findings provide useful
cues for the construction of highly stable, mixed monolayer
the stability of glyco-AuNPs. a) The stability of glyco-AuNPs in various NaCl
solutions. Changes in the SPR band intensities [(AÀA
0 0
), A =initial absorb-
ance, and A=absorbance after 8 h] of different Lac-AuNPs (1–AuNP to 5–
AuNP) in the presence of varying concentrations of NaCl at 620 nm after 8 h
at room temperature. b) The UV/Vis absorption spectra of 5–AuNP treated
with BSA at concentrations of 0, 0.02, 2, and 200 mm in PBS after 8 h incuba-
H
tion. The inset shows the corresponding D of 5–AuNP in the presence of
BSA at each indicated concentration.
Chem. Eur. J. 2015, 21, 3956 – 3967
www.chemeurj.org
3959
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
protected glyco-AuNPs for bioanalytical assays (see below) sim-
induce the aggregation of glyco-AuNPs, which is reflected in
the color change of the glyco-AuNP from red to purple or
blue, and can be seen easily with the naked eye.
[
16]
ilar to that of PEGylated AuNPs.
In addition to enhancing the colloidal stability of glyco-
AuNPs in high physiological salt conditions (NaCl, 150 mm), the
presence of 12 led to little or no nonspecific protein adsorp-
tion. As shown in Figure 3b, no apparent changes in the sur-
face plasmon absorbance and hydrodynamic diameter (DH)
were observed upon treatment of 5–AuNP with bovine serum
albumin (BSA; 0.02 to 200 mm), indicating that the mixed mon-
olayer renders the AuNPs inert to nonspecific adsorption of
proteins in solution. Although 2–AuNP obviated nonspecific
binding to BSA (see the Supporting Information, FigureS4), 1–
AuNP, which lacks the 12 monolayer, exhibited nonspecific BSA
adsorption as indicated by the UV/Vis spectrum and DLS meas-
urements (see the Supporting Information, FigureS5). Never-
theless, the presence of 12 is crucial because it not only im-
parts steric hindrance to proteins that could be nonspecifically
adsorbed on the nanoparticle surface, but also provides
As shown in Figure 4a, lectins specifically recognize and
bind to their cognate carbohydrate ligands present on the sur-
[33]
a good nanoparticle packing density in aqueous solution.
Furthermore, glyco-AuNPs containing other glycan structures
with linker 12, such as 6–AuNP and 7–AuNP, also exhibited no
change in the surface plasmon absorption band; consequently,
the nonspecific adsorptions were insignificant (see the Sup-
porting Information, FigureS6). These results indicate that
a 30% monolayer of 12 in glyco-AuNPs impedes nonspecific
protein attachments, in good agreement with an earlier report
that an ethylene glycol monolayer prevents the nonspecific
[
34]
binding of proteins such as BSA to AuNPs.
Recognition of glyco-AuNPs by lectins
Lectins have been employed widely as a model system for in-
vestigating how proteins recognize glycans on a molecular
level through selective carbohydrate–protein interactions. The
recognition between lectins and glycans essentially occurs
through multiple low-affinity interactions. Glyco-AuNPs have
potential as ideal scaffolds for presenting multiple carbohy-
drates, and are similar in size to many biomolecules. For fur-
ther investigation of the specific binding interactions of lectins
and carbohydrates, a representative panel of seven different
lectins was selected. Table 1 summarizes the synthetic AuNPs
used in this study and the lectin specificity to the correspond-
ing glycan structures present on the synthesized glyco-AuNPs.
We describe here a sensitive plate-based colorimetric assay for
rapid profiling of the presence of lectins. In this assay, lectins
Figure 4. Glyco-AuNP-based colorimetric lectin assay. a) Plate-based colori-
metric assay for lectins with different glyco-AuNPs (5–AuNP to 11–AuNP).
b) AuNP-based colorimetric detection of RCA120. The inset shows the linear
range from 0 to 10 nm RCA120 concentration, which was used to calculate
the limit of detection (LOD) for RCA120
.
faces of the glyco-AuNPs. Of these plant lectins, Con A, MAL,
and SNA displayed good affinities to 6–AuNP, 9–AuNP, and 10–
AuNP, respectively. As expected, the fungal lectin AAL, which
recognizes the l-fucose residue, exhibited preferential binding
to 8–AuNP. The bacterial lectin PA-IL, which recognizes only
Gb3 glycans, only interacted with 11–AuNP. Despite possessing
a nominal binding specificity for Lac (Galb1,4Glc), RCA120 also
binds nonreducing Gal-modified Lac. The results of the colori-
metric assays revealed strong RCA120 binding to nonreducing
k
[42]
terminal Gal, such as that found in the P antigen (11–AuNP)
[43]
and NeuAca2,6-Gal
(10–AuNP), but not NeuAca2,3-Gal, in
Table 1. The lectins, their carbohydrate ligands, and the corresponding
glyco-AuNP ligands.
good agreement with previous reports. Conversely, the
GlcNAc-specific lectin WGA exhibited preferential binding to
Glyco-AuNP
Lectin
Carbohydrate ligand recognized
[Ref.]
7
–AuNP, as predicted, but also interacted with the NeuAca2,3-
5–AuNP
6–AuNP
7–AuNP
8–AuNP
9–AuNP
10–AuNP
RCA120
Con A
WGA
AAL
MAL
SNA
Lac: Galb1,4Glc
a-d-Man
b-d-GlcNAc
[35]
[36]
[37]
[38]
[39]
[40]
[41]
Gal-terminating structure (9–AuNP), also in accordance with lit-
[44]
erature data.
Addition of the free carbohydrates Lac
a-l-Fuc
(10 mm), Man-a-OMe (100 mm), or GlcNAc (500 mm) in excess
to aggregated AuNP solutions containing RCA120 and 5–AuNP,
Con A and 6–AuNP, or WGA and 7–AuNP, respectively, promot-
ed complete redispersion of the aggregates along with recov-
GM3: NeuAca2,3Galb1,4Glc
NeuAca2,6Galb1,4Glc
Gb3: Gala1,4Galb1,4Glc
11–AuNP
PA-IL
Chem. Eur. J. 2015, 21, 3956 – 3967
www.chemeurj.org
3960
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
3
ery of the original surface plasmon band (data not shown).
These results indicate that the aggregation was specifically
caused by carbohydrate–lectin interactions. Furthermore, recip-
rocal experiments between glyco-AuNPs (5–AuNP, 6–AuNP,
and 7–AuNP) and lectins such as RCA120, Con A, and WGA re-
vealed no nonspecific binding between carbohydrates and lec-
tins. These results clearly indicate the potential for the use of
glyco-AuNPs in the detection of target lectins.
ure S9), which is 2.2ꢂ10 times more sensitive than a Gal-func-
[24]
tionalized silica nanoparticle in a DLS assay.
The LOD for
Con A using 6-AuNP was 16 pm, which is approximately 9.7ꢂ
2
[24]
10 times higher than a similar DLS measurement. The sensi-
tivity for WGA was as low as 12 pm if 7–AuNP was used. Nota-
bly, most colorimetric sensors based on nanoparticle aggrega-
[18,23,24,45]
tion exhibit LODs greater than 1.5 nm;
LODs are improved to low picomolar levels by employing DLS.
however, the
Determination of the detection sensitivity of different lec-
tins by UV/Vis spectroscopy and DLS
Glyconanoparticle-based oriented antibody microarray for
lectin sensing (GOAL assay)
Next, we investigated the detection sensitivity of the AuNP-
based colorimetric bioassay for individual lectins by using UV/
Vis spectroscopy (Table 2). The assays were performed by treat-
We previously developed a covalent and site-specifically immo-
bilized Fc-fused lectin microarray, and demonstrated that the
oriented, immobilized lectin retained higher lectin activity
[46]
compared with random immobilization. The use of BA as an
affinity head group to direct the orientation of the antibody
has been extended to immobilize them on solid supports such
Table 2. LODs of lectins estimated by different methods using glyco-
AuNPs.
[47]
as magnetic nanoparticles. Furthermore, we previously used
[
b]
[c]
a typical “sandwich” approach for the detection of the B subu-
nit of Slt-I bound to a randomly immobilized antibody microar-
ray surface, achieving a detection sensitivity of around
Glyco-AuNP
Lectin
Colorimetric
DLS [pm]
GOAL assay
[
a]
[pm]
[
17]
[24]
5–AuNP
6–AuNP
7–AuNP
RCA120
Con A
WGA
300 (7 nm)
3 (6.6 nm)
16 (15 nm)
1
5
1
[25]
[
[
23]
[24]
430 (6.2 nm)
500 (1.5 nm)
70 nm. Therefore, to develop a glyco-AuNP-based immuno-
sensing platform, we hypothesized that an oriented antibody
microarray could be constructed using BA-derivative 26, which
would preserve the activity of the immobilized antibody, re-
45]
12
The LOD values were calculated as follows: [a] LOD=signal of the blank
3*standard deviation (SD). [b] LOD=3*SD of the blank/slope of the cal-
ibration curve. [c] Determined as the lowest concentration that produced
a detectable signal after silver enhancement. The corresponding LOD
values reported in the literature are shown in parentheses.
+
[48]
sulting in an improved detection sensitivity. Because of its
hydrophilic nature, the tri(ethylene glycol) spacer in 26 pre-
vented nonspecific adsorption and minimized detrimental in-
teractions between the attached protein and the solid sur-
[49]
face.
A homogeneous interaction of the bound antigen
ing 5–AuNP (3 nm) with a range of concentrations of RCA120 (0,
could also be achieved by simultaneous incubation of the
high-affinity glyco-AuNPs on the microarray surface. These
studies would validate the utility of the multivalent effect for
designing a glyconanoparticle-based oriented antibody micro-
array for lectin sensing (GOAL assay; Figure 1). For implemen-
tation of the GOAL assay, the effect of the size of the nanopar-
ticle, which serves as the seed for silver enhancement, on the
visualization of in situ agglutination was first investigated
(Figure 5). Lectin-induced in situ AuNP agglutination with
visual detection is an easy-to-use method that avoids the use
of sophisticated instruments for the elucidation of carbohy-
drate-mediated interactions.
0.1, 0.5, 1, 2.5, 5, 7.5, 10, 20, 30, and 40 nm), a potential surro-
gate for the bioterrorism agent ricin. The aggregation of 5-
AuNPs was indicated by the color change of the AuNPs from
red to blue (see the Supporting Information, FigureS7) and a si-
multaneous change in the absorption band in the visible
region of the electromagnetic spectrum (Figure 4b). As illus-
trated in Figure 4b, the plot of the change in the SPR band in-
tensity at 620 nm versus various concentrations of RCA120 re-
veals a linear concentration-dependent aggregation from 0 to
10 nm, with a limit of detection (LOD) for RCA120 of 300 pm
(
Table 2), which is 23 times higher than that estimated from
[
17]
optimally presented Gal-coated AuNPs. The higher detection
sensitivity of 5–AuNP may be attributed to its superior stability
and optimal presentation of Lac head groups in an ordered
monolayer. These nanoparticles enable more efficient intimate
As a demonstration of the feasibility of the GOAL assay, the
anti-RAC (A-chain) antibody (10 mm) was fabricated on a 26-
coated surface (50 mm). Following blocking with dextran
(100 mm containing 1% BSA), the slide was incubated with a so-
lution of RCA120 (20 nm) and 5–AuNP (5 nm) for 8 h at room
temperature (Figure 5a). As a control, the anti-RAC antibody
(10 mm) was immobilized randomly on a Nexterion H slide by
amide bond formation, followed by simultaneous incubation
with RCA120 and 5–AuNP, as described above (Figure 5b).
RCA120 is a heterodimeric protein consisting of two A–B subu-
[11,18]
contacts with the carbohydrate-binding sites of the lectin.
On the basis of this colorimetric bioassay (Table 2), subnano-
molar detection sensitivities (or LODs) for Con A (430 pm) and
WGA (500 pm) were also achieved using 6–AuNP and 7–AuNP
within a linear range from 0.1 to 10 nm and 0.5 to 7.5 nm for
Con A and WGA, respectively (see the Supporting Information,
FigureS8).
[35]
nits and a Lac-binding site on each B subunit. Therefore, the
lectin can act as a crosslinker. Antibodies on the surface cap-
tured the A chain(s) of RCA120, and the remaining two Lac-bind-
ing B chains and RCA120 in solution aggregated simultaneously
The AuNP-based colorimetric sensing method was also com-
pared with the results obtained using DLS (Table 2). The LOD
for RCA120 was 3 pm (see the Supporting Information, Fig-
Chem. Eur. J. 2015, 21, 3956 – 3967
www.chemeurj.org
3961
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
further immunosensing experiments for LOD measurements
(
see below).
For the determination of the LOD of RCA120, the array slide
was exposed to 100 mL of solutions of different concentrations
of RCA120 (1–10 nm, final concentration) and 5–AuNP (5 nm)
with gentle agitation at 30 rpm for 2 h. As shown in Figure 6a,
Figure 5. Schematic depiction of the GOAL assay for lectin detection. The
capture antibody (10 mm) that recognizes lectin, RCA120, was fabricated by
a) oriented immobilization through boronate formation using a BA deriva-
tive (26), and b) random amide bond formation on an NHS-activated glass
slide. The presence of the bound antigen was detected by simultaneous in-
cubation with RCA120 (20 nm) and a high-affinity glyco-AuNP, 5–AuNP (5 nm),
followed by silver enhancement.
Figure 6. In situ lectin-induced glyco-AuNP agglutination immunoassay for
lectin detection. Silver-enhanced images of slides after incubation with
a) RCA120 (1, 5, and 10 nm) with 5–AuNP; b) Con A (1, 5, and 10 nm) with 6–
AuNP; or c) WGA (1, 5, and 10 nm) with 7–AuNP for 2 h. A representative
control slide without RCA120 and exposed to 5–AuNP is also shown. The im-
munoassay was performed with lectin solutions in PBS buffer. All images
were acquired with a flatbed scanner.
with 5–AuNPs. Following a final wash, the array slides were ex-
posed to the silver amplification solution for 14 min to darken
the aggregated AuNPs to enable visualization of the spots
with the naked eye. As evident from the silver-enhanced
image in Figure 5a, the antibody microarray immobilized in an
oriented manner produced a homogeneous array definition,
whereas the microarrays fabricated by random antibody immo-
bilization furnished a very weak signal after silver enhance-
ment (Figure 5b). The appearance of darker spots in the
former method may be because of the increased capture of
AuNPs on the slide surface. This increased capture may occur
because the surface-conjugated BAs direct the formation of
cyclic boronate diesters with the diol groups of glycans at the
Fc region of the antibody, thereby leaving the Fab region ex-
spots were clearly identified at concentrations as low as 1 nm
RCA120 after silver enhancement. No spots were observed on
the control slide without RCA120, demonstrating that neither
nonspecific adherence of the AuNPs to the immobilized anti-
body nor nonspecific silver deposition on the bare glass oc-
curred. Importantly, additional control experiments using
glyco-AuNPs (1–AuNP to 3–AuNP) lacking a mixed monolayer
revealed nonspecific adsorption on the slide surface or on the
antibodies (see the Supporting Information, FigureS11). On the
basis of the visible spots formed by specific agglutination and
silver precipitation, a detection sensitivity of 1 nm was ach-
ieved. This sensitivity is approximately six times greater than
that of reported literature methods based on UV/Vis spectros-
[50]
posed to the interface for enhanced antigen capture and
leading to a greater number of AuNP seeds available for nucle-
ation. The results strongly suggest that the orientation of anti-
bodies on the microarray is significant for the enhancement of
the detection sensitivity.
[17,24]
copy and DLS (see Table 2).
We suggest that the observed sensitivity comes from the
control of the orientation of the antibodies on the BA surface,
which increases the area available for antigen binding, result-
ing in stable complexes with a high antigen/antibody ratio.
AuNPs can undergo a favorable (homogeneous) interaction
with RCA120 binding sites in solution owing to the minimal sur-
face steric hindrance, which induces increased AuNP aggrega-
tion during in situ agglutination. Notably, the shortest distance
LODs of different lectins in the GOAL assay
After establishing that orientation is critical for providing im-
proved antibody activity during in situ agglutination, we then
used the GOAL assay to estimate the detection sensitivities of
different antigens. For further optimization of the detection
sensitivity of the GOAL assay, the effects of incubation time (1–
[51]
8
h) and AuNP concentration (0.5–5 nm) were investigated (see
between two binding sites in RCA120 is approximately 10 nm.
the Supporting Information, FigureS10). For a fixed concentra-
tion of RCA120 (10 nm), a 5 nm concentration of 5–AuNP and
a 2 h incubation time produced an acceptable level of array
definition. Therefore, these optimal conditions were used for
Therefore, the glyco-AuNP 5–AuNP (15 nm core diameter) used
in the present study, which has a circumference of around
47 nm, may permit complete insertion of Lac into the binding
pocket of RCA120. This insertion leads to the formation of many
Chem. Eur. J. 2015, 21, 3956 – 3967
www.chemeurj.org
3962
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
layers on the microarray surface similar to the previously de-
scribed binding of RCA using AuNPs with 70% Gal cover-
AuNP, and 7–AuNP. Silver enhancement was then performed to
enlarge and darken the captured nanoparticles.
120
[
17]
age.
Upon examination of Con A and WGA, similarly low
As shown in Figure 7b–d, each lectin produced distinct
spots at appropriate regions where its cognate capture Ab was
immobilized on the surface. In addition, Figure 7e clearly indi-
cates the spots resulting from a solution containing all three
lectins. Minimal nonspecific binding was observed. The only
lectin that exhibited strong cross-reactivity was RCA120. When
detected as part of a multiantigen immunosensing assay,
RCA120 recognized and bound both anti-WGA and anti-Con A
antibodies (Figure 7b). This cross-reactivity may be caused by
glycosylation of the anti-WGA and anti-Con A antibodies by
additional NeuAca2,6Gal- or terminal Gal-containing glycan
structures accessible for binding to lectin RCA120, which can
function as a glycoprotein.
nanomolar detection sensitivities of 5 nm and 1 nm, respective-
ly, were established (Figure 6b,c and Table 2). Nevertheless,
these results demonstrate that selective in situ agglutination of
AuNPs on microarrays can be applied widely for the detection
of carbohydrate-binding proteins.
Although a low picomolar-level lectin-sensing platform
could be realized on the basis of DLS measurements (Table 2),
we have determined that this method is not capable of detect-
ing analytes in the context of environmental settings, such as
in samples from lake water. The strong background scattering
contributes more significantly to the absolute scattering light
intensity despite prior passage of the sample through
a 0.45 mm filter (see the Supporting Information, FigureS12a).
Therefore, the DLS-based assay is unsuitable for samples in
such conditions. The SPR band of 5–AuNP in lake water re- Conclusions
mained unchanged at 525 nm, suggesting that 5–AuNPs are
In conclusion, we have developed highly stable, mixed-mono-
stable and do not aggregate even in the presence of other mi-
croparticles (FigureS12b). Notably, our GOAL assay was still
able to detect 2.5 nm RCA120 under these conditions (Fig-
ure S12c). These findings highlight the potential of the GOAL
assay for the sensing of analytes under diverse environmental
stress conditions.
layer-protected glyco-AuNPs and used them as model nanocar-
riers to study the specific biomolecular recognition between
lectins and carbohydrates. These glyco-AuNPs exhibited com-
plete resistance to nonspecific protein adherence in solution
and on solid supports. Specific binding interactions for differ-
ent carbohydrates and lectins, such as RCA120 with Lac, Con A
with Man, and WGA with GlcNAc, were established on the
basis of colorimetric and DLS measurements. The lectin ana-
lytes were detected individually by utilizing in situ glyco-
AuNPs agglutination-based immunosensing on oriented anti-
body microarrays with the naked eye. The method is applica-
ble regardless of the lectins and type of saccharides present
on the nanoparticles. We have also described the ability of
glyco-AuNP-based immunosensing to detect multiple lectins in
samples with low nonspecific cross-reactivity. Owing to the
ease of detection and the miniature format of this array tech-
nology, we anticipate that the glyco-AuNP-based immunosens-
ing strategy is amenable to the production of chips for appli-
cations in bioassays and biorecognition.
Simultaneous detection of multiple antigens
As a demonstration of the capability of this newly developed
method for multiple lectin detection, the GOAL assay was used
to detect multiple lectins in the same sample. Capture Ab
spots (12 replicate spots per Ab) were printed in a 2ꢂ6 array
format on a 26-functionalized glass slide to detect the pres-
ence of RCA120 (20 nm), Con A (20 nm), and WGA (20 nm) on
a single microarray. Figure 7b–e shows the representative
silver-enhanced images obtained when these three lectins
were presented as a single lectin (Figure 7b–d) or as a mixture
of multiple lectins (Figure 7e). Detection was achieved by ex-
posure of each microarray to the following glyco-AuNPs
(5 nm), each aimed at one of the target lectins; 5–AuNP, 6–
Experimental Section
General experimental methods
See the Supporting Information for details of general experimental
methods.
Syntheses
The syntheses of thiolated lactoside 3, linker 12, and amino-termi-
nated tri(ethylene glycol)-linked boronic acid (26) are described in
the Supporting Information. Azido-terminated tri(ethylene glycol)-
linked saccharides 14, 15, 16, 17, and linker 13 were
obtained by following literature procedures as described previous-
Figure 7. Simultaneous detection of multiple lectins using the GOAL assay.
a) A microarray was patterned on each column with twelve replicate capture
Ab spots (anti-RAC (A-chain) pAb, anti-Con A pAb, and anti-WGA pAb) corre-
sponding to three lectins as depicted in the schematic. Three samples con-
taining a single lectin (b–d) and one sample containing a mixture of three
lectins (e) were incubated for 2 h. The detection of bound lectin (20 nm) was
visualized with the corresponding glyco-AuNPs followed by silver enhance-
ment. Representative photographs of the microarray after silver staining for
[52]
[53]
[54]
[55]
[26]
[
26]
k
ly. The thiolated a-2,3-sialyllactoside (9) and azido-terminated P
[56]
trisaccharide (19) were obtained from chemoenzymatic methods
we described recently. a-2,6-Sialyllactoside derivative 18 was syn-
thesized (see the Supporting Information) in an analogous manner,
[26]
14 min: b) RCA120, c) Con A, d) WGA, and e) three lectins.
as reported by us.
Chem. Eur. J. 2015, 21, 3956 – 3967 www.chemeurj.org
3963
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Synthesis of compound 6
4.7, 10.0 Hz, 1H, H5), 3.68–3.75 (m, 1H, -OCH
2
-), 3.57–3.67 (m, 10H,
-
3
7
OCH -), 3.47–3.57 (m, 6H, -OCH -), 3.35 (t, J=5.6 Hz, 2H, -NHCH -),
.26–3.32 (m, 4H, -NHCH - overlapped with CD OD), 2.85 (t, J=
2 3
2
2
2
Palladium on carbon (Pd/C; 100.0 mg; 10% Pd content) was sus-
pended in
0
5
a
solution of azido a-mannoside 15 (500.0 mg,
.99 mmol) in MeOH (20 mL) in the presence of AcOH (0.3 mL,
.24 mmol). The flask was purged with H , and a hydrogen balloon
.3 Hz, 2H, -CH S-), 2.29 (s, 3H, -(C=O)CH ), 2.18 (t, J=7.5 Hz, 2H,
2
3
-
(C=O)CH -), 2.04 (s, 3H, -(C=O)CH ), 2.00 (s, 3H, -(C=O)CH ), 1.97 (s,
2
3
3
2
3
1
1
7
3
2
H, -(C=O)CH ), 1.90 (s, 3H, -(C=O)CH ), 1.49–1.65 (m, 4H, -CH -),
3 3 2
was attached. The solution was stirred for 3 h at ambient tempera-
ture, and the catalyst was filtered off through a pad of Celite. The
Celite bed was washed with MeOH, and the combined filtrates
were concentrated under reduced pressure to provide the amine,
which was used in the next step without further purification. Trie-
thylamine (1.4 mL, 10.06 mmol) and linker 13 (660 mg, 1.2 mmol)
were added to a solution of amine in dichloromethane (30 mL) at
À208C. The resulting mixture was warmed gradually to room tem-
perature and stirred for another 6 h until TLC indicated complete
disappearance of the starting material. The reaction mixture was
partitioned into dichloromethane and water. The organic layer was
13
.24–1.42 ppm (m, 12H, -CH -); C NMR (100 MHz, CDCl ): d=
2
3
95.9, 173.6, 170.7, 170.7, 170.5, 169.2, 158.6, 101.0, 72.9, 71.4, 70.6,
0.5, 70.4, 70.3, 70.1, 69.9 (ꢂ2), 69.7, 68.6 (ꢂ2), 62.0, 53.9, 39.8,
9.6, 39.0, 36.3, 30.4, 29.2, 29.2 (ꢂ3), 29.0, 28.9, 28.8, 28.5, 25.5,
2.8, 20.5 (ꢂ2), 20.4 ppm; HRMS (ESI): m/z calcd for C H N O SNa
40
70
4
16
+
[
M+Na] : 917.4405; found: 917.4407. Compound 7 was prepared
using the procedures described above, starting with 22 (710 mg,
.79 mmol). Purification by silica gel column chromatography yield-
ed 7 (334 mg, 0.46 mmol, 58%). R =0.42 (MeOH/DCM=1:5);
0
f
29
1
[a] = +11.72 (c=1.0 in MeOH); H NMR (400 MHz, CD OD): d=
D
3
4
3
.45 (d, J=8.3, 1H, H1), 3.94 (ddd, J=3.8, 3.8, 11.2 Hz, 1H, -OCH -),
2
washed successively with saturated NaHCO (aq.), 1m aqueous HCl,
3
.86 (dd, J=2.1, 12.1 Hz, 1H, H6a), 3.56–3.73 (m, 14H, H2, H3, H6b,
and brine, then dried over MgSO , filtered, and concentrated in va-
4
-OCH -), 3.42 (dd, J=8.2, 10.2 Hz, 1H, H4), 3.22–3.37 (m, 7H, H5,
2
cuo. Purification by silica gel column chromatography yielded 21
-NHCH - overlapped with CD OD), 2.47 (t, J=7.3 Hz, 2H, -CH S-),
2
3
2
(
1
613 mg, 0.68 mmol, 69% over two steps). R =0.42 (MeOH/DCM=
f
2
1
.17 (t, J=7.7 Hz, 2H, -(C=O)CH -), 1.97 (s, 3H, -(C=O)CH ), 1.52–
3
0
1
2
3
13
:15); [a] = +21.83 (c=1 in MeOH); H NMR (400 MHz, CDCl3):
D
.64 (m, 4H, -CH -), 1.25–1.44 ppm (m, 12H, -CH -); C NMR
2
2
d=6.22 (br s, 1H, -NH-), 5.17–5.39 (m, 5H, H2, H3, H4,-NH-), 4.86
(100 MHz, CD OD): d=176.2, 173.6, 160.9, 102.7, 77.9, 76.0, 72.0,
3
(
d, J=1.6 Hz, 1H, H1), 4.25 (dd, J=5.0, 12.4 Hz, 1H, H6a), 4.03–4.14
7
3
2
7
1.4 (ꢂ3), 71.4, 71.2 (ꢂ2), 71.2, 70.5, 69.8, 62.7, 57.1, 40.8 (ꢂ2), 40.2,
(
m, 2H, H6b, H5), 3.76–3.84 (m, 1H, -OCH -), 3.48–3.73 (m, 17H,
2
7.0, 35.1, 30.5, 30.5, 30.3, 30.2, 30.1, 29.3, 26.9, 25.1, 24.9,
-
OCH -), 3.29–3.46 (m, 6H, -NHCH -), 2.82 (t, J=7.4 Hz, 2H, -CH S-),
+
2
2
2
3.1 ppm; HRMS (ESI): m/z calcd for C H N O SNa [M+Na] :
32
62
4
12
2
3
1
1
1
6
2
.29 (s, 3H, -(C=O)CH ), 2.15 (t, J=7.4 Hz, 2H, -(C=O)CH -), 2.13 (s,
3
2
49.3983; found: 749.3958.
H, -(C=O)CH ), 2.08 (s, 3H, -(C=O)CH ), 2.01 (s, 3H, -(C=O)CH ),
3
3
3
.95 (s, 3H, -(C=O)CH ), 1.46–1.66 (m, 4H, -CH -), 1.15–1.38 ppm (m,
3
2
1
3
2H, -CH -); C NMR (100 MHz, CDCl ): d=195.9, 173.5, 170.6,
2
3
Synthesis of compound 8
69.9, 169.5, 158.4, 97.4, 70.6, 70.5, 70.4, 70.2, 70.0, 69.9, 69.7(ꢂ2),
9.3, 68.9, 68.1, 7.2, 65.8, 62.2, 39.8(ꢂ2), 39.0, 36.4, 30.4, 29.2,
The fully protected thiolated fucoside 23 was prepared using the
procedures described in the synthesis of compound 21, starting
with 17 (500.0 mg, 0.99 mmol) and linker 13 (683 mg, 1.23 mmol).
Purification by silica gel column chromatography produced 23
9.2(ꢂ2), 29.1, 29.0, 28.9, 28.8, 28.5, 25.5, 20.7, 20.6, 20.5 ppm;
+
HRMS (ESI): m/z calcd for C H N O S [M+H] : 896.4426; found:
40
70
3
17
8
96.4459. The fully protected thiolated mannoside 21 (450 mg,
0
.50 mmol) was dissolved in dry MeOH (10 mL), and NaOMe
(697 mg, 0.83 mmol, 74% over 2 steps). R =0.42 (MeOH/DCM=
f
(
40.5 mg, 0.75 mmol) was added. The resulting mixture was stirred
29
1
1
:20); [a] = +56.38 (c=1.0 in MeOH); H NMR (400 MHz, CD OD):
D
3
at room temperature for 1 h and neutralized with Amberlite ion-ex-
change resin IR 120. After filtration and concentration, flash silica
gel column chromatography provided thiolated mannoside deriva-
d=5.33 (dd, J=3.4, 10.9 Hz, 1H, H3), 5.26 (br d, J=3.4 Hz, 1H, H4),
.07 (d, J=3.6 Hz, 1H, H1), 5.03 (dd, J=3.6, 10.9 Hz, 1H, H2), 4.27
br q, J=6.5 Hz, 1H, H5), 3.76–3.84 (m, 1H, -OCH -), 3.57–3.72 (m,
5
(
1
-
2
tive 6 (220 mg, 0.32 mmol, 64%). R =0.18 (MeOH/DCM=1:9);
f
1H, -OCH -), 3.45–3.56 (m, 6H, -OCH -), 3.34 (t, J=5.4 Hz, 2H,
3
0
1
2
2
[
a] +16.29 (c=1 in MeOH); H NMR (400 MHz, CD OD): d=4.81
D
3
NHCH -), 3.25–3.31 (m, 4H, -NHCH - overlapped with CD OD), 2.84
2
2
3
(
3
d, J=1.5, 1H, H1), 3.79–3.87 (m, 3H), 3.56–3.74 (m, 15H), 3.48–
.56 (m, 6H), 3.35 (t, J=5.6 Hz, 2H, -NHCH -), 3.26–3.32 (m, 4H,
(t, J=7.2 Hz, 2H, -CH S-), 2.28 (s, 3H, -(C=O)CH ), 2.17 (t, J=7.6 Hz,
2
3
2
2
1
1
H, -(C=O)CH -), 2.13 (s, 3H, -(C=O)CH ), 2.04 (s, 3H, -(C=O)CH ),
2 3 3
-NHCH - overlapped with CD OD), 2.38 (t, J=7.4 Hz, 2H, -CH S-),
2 3 2
.94 (s, 3H, -(C=O)CH ), 1.39–1.55 (m, 4H, -CH -), 1.16–1.32 (m,
3
2
13
2
1
1
6
3
.18 (t, J=7.9 Hz, 2H, -(C=O)CH -), 1.53–1.65 (m, 4H, -CH -), 1.26–
2
2
2H, -CH -), 1.02 ppm (d, J=6.5 Hz, 3H, -CH ); C NMR (100 MHz,
13
2
3
.44 ppm (m, 12H, -CH -); C NMR (100 MHz, CD OD): d=176.7,
2
3
CD OD): d=197.2 176.2, 176.1, 172.1, 171.8, 171.5, 160.8, 97.4, 72.4,
3
61.1, 101.7, 74.6, 72.5, 72.1, 71.6, 71.4, 71.3 (ꢂ3), 71.3, 71.2, 70.5,
8.6, 67.6, 62.9, 41.3, 41.2, 40.5, 36.8, 35.2, 30.6, 30.5, 30.4, 30.2,
0.2, 29.4, 27.0, 25.0 ppm; HRMS (ESI): m/z calcd for C H N O S
7
4
2
1.6, 71.5, 71.5, 71.3 (ꢂ3), 71.2, 70.5, 69.3 (ꢂ2), 68.4, 65.5, 40.9,
0.8, 40.2, 37.0, 30.7, 30.6, 30.4 (ꢂ4), 30.3, 30.2, 30.1, 29.8, 29.7,
30
58
3
12
6.9, 20.7, 20.7, 20.5, 16.2 ppm; HRMS (ESI): m/z calcd for
À
[MÀH] : 684.3741; found: 684.3704.
+
C H N O SNa [M+Na] : 860.4191; found: 860.4151. Compound 8
38
67
3
15
was prepared using the procedure described for 6, starting with
3 (340 mg, 0.41 mmol). Purification by flash silica gel column
chromatography afforded thiolated fucoside (160.0 mg,
0.24 mmol, 59%). R =0.55 (MeOH/DCM=1:9). [a] =51.23 (c=1.0
2
Synthesis of compound 7
8
2
D
9
The fully protected thioacetate 22 was prepared using the proce-
dures described in the synthesis of compound 21, starting with 16
f
1
in MeOH); H NMR (400 MHz, CD OD): d=4.85 (d, J=3.1, 1H, H1),
3
(
500 mg, 0.99 mmol) and linker 13 (660.0 mg, 1.20 mmol). Purifica-
tion by silica gel column chromatography afforded 22 (712 mg,
.80 mmol, 80% over two steps). R =0.37 (MeOH/DCM=1:15);
4.04 (br q, J=6.6 Hz, 1H, H5), 3.84 (ddd, J=4.0, 4.0, 10.5 Hz, 1H,
-OCH -), 3.62–3.80 (m, 14H), 3.52–3.61 (m, 6H), 3.39 (t, J=5.4 Hz,
2
0
2H, -NHCH -), 3.34 (t, J=5.3 Hz, 4H, -NHCH - overlapped with
f
2
2
2
9
1
[a] = +10.07 (c=1.0 in MeOH); H NMR (400 MHz, CD OD): d=
CD OD), 2.52 (t, J=7.1 Hz, 2H, -CH S-), 2.08 (t, J=7.7 Hz, 2H,
3 2
D
3
5
.20 (dd, J=10.0, 10.0 Hz, 1H, H3), 4.97 (dd, J=10.0, 10.0 Hz, 1H,
-(C=O)CH -), 1.56–1.76 (m, 4H, -CH -), 1.29–1.51 (m, 12H, -CH -),
2 2 2
13
H4), 4.70 (d, J=8.4 Hz, 1H, H1), 4.27 (dd, J=4.7, 12.4 Hz, 1H, H6a),
1.25 ppm (d, J=6.6 Hz, 3H, -CH3); C NMR (100 MHz, CD OD): d=
3
4
1
.27 (dd, J=2.4, 12.4 Hz, 1H, H6b), 3.92 (ddd, J=4.2, 4.2, 11.1 Hz,
176.3, 161.0, 100.6, 73.5, 71.7, 71.6, 71.5, 71.5, 71.3, 71.2 (ꢂ3), 70.6,
70.0, 68.2, 67.6, 40.9, 40.8, 40.2, 37.0, 35.2, 30.6, 30.5, 30.4, 30.2,
H, -OCH -), 3.85 (dd, J=8.4, 10.0 Hz, 1H, H2), 3.79 (ddd, J=2.4,
2
Chem. Eur. J. 2015, 21, 3956 – 3967
www.chemeurj.org
3964
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
3
0.2, 29.4, 27.0, 25.0, 16.7 ppm; HRMS (ESI): m/z calcd for
trisaccharide derivative 11 (199 mg, 0.20 mmol, 60% over three
+
29
C H N O SNa [M+Na] : 692.3768; found: 692.3795.
steps). R =0.45 (EtOAc/MeOH/H O/AcOH=3/2.5/0.5/0.5); [a] =
3
0
59
3
11
f
2
D
1
+
30.11 (c=1.0 in H O); H NMR (600 MHz, D O plus one drop
2 2
CD OD): d=4.87 (d, J=3.6 Hz, a-Gal-H1), 4.40 (d, J=7.7 Hz, 1H),
3
Synthesis of compound 10
4
.37 (d, J=7.9 Hz, 1H), 4.24 (br t, J=6.3 Hz, 1H, a-Gal-H5), 3.97
The sialyllactoside 24 was prepared according to the procedure de-
scribed for 6, starting with 18 (100 mg, 0.13 mmol) and linker 13
(ddd, J=4.2, 4.2, 11.3 Hz, 1H), 3.95 (br d, J=3.1 Hz), 3.92 (dd, J=
1.0, 3.0 Hz), 3.88 (dd, J=2.1, 12.3 Hz), 3.84 (dd, J=2.1, 12.3 Hz),
(
90.0 mg, 0.16 mmol), with the exception that hydrogenation was
3.42–3.81 (m, 29H), 3.30 (t, J=5.4 Hz, 2H, -NHCH
5H, -NHCH -, Glc-H2), 2.62 & 2.44 (t, 2H, J=7.2 & 7.1 Hz, -CH
from disulfide dimer & monomer, respectively), 2.15 (t, J=7.4 Hz,
2H, -(C=O)CH -), 1.46–1.55(m, 4H, -CH -), 1.26–1.34(m, 2H,
-CH -),1.15–1.26 ppm (m, 10H, -CH -); C NMR (100 MHz, D O plus
one drop CD OD): d=176.4, 161.0, 104.2, 103.1, 101.3, 79.8, 78.4,
2
-), 3.21–3.26 (m,
performed in the absence of AcOH. Purification by size exclusion
column chromatography on Bio-Gel P-2 afforded sialyllactoside 24
(
2
2
S-
108.0 mg, 0.09 mmol, 74% over two steps). R =0.29 (EtOAc/
2
2
f
29
13
MeOH/H O/AcOH=3/2.5/0.5/0.5); [a] =À7.74 (c=2.4 in MeOH);
2
2
2
2
D
1
H NMR (400 MHz, D O): d=4.41 (d, J=8.0, 1H, Gal-H1), 4.33 (d, J=
3
2
7
3
.8, 1H, Glc-H1), 3.92–4.02 (m, 1H, -OCH -), 3.39–3.91 (m, 32H),
76.3, 76.3, 75.7, 75.3, 73.9, 73.2, 71.9, 71.8, 70.9, 70.8, 70.7, 70.6 (ꢂ
2), 70.6, 70.4, 70.1, 70.1, 70.0, 69.5, 61.6, 61.3, 61.1, 40.4 (ꢂ2), 39.8,
2
.18–3.32 (m, 7H, -NHCH -, Sia-H5), 2.76 (t, J=7.3, 2H, -CH S-), 2.61
2
2
(
7
1
dd, J=4.5, 12.2, 1H, Sia-H3a), 2.25 (s, 3H, -(C=O)CH ), 2.14 (t, J=
36.8, 34.7, 30.4 (ꢂ2), 30.2, 29.9 (ꢂ2), 29.2, 26.6, 25.2 ppm; HRMS
3
+
.3, 2H, -(C=O)CH -), 1.93 (s, 3H, -(C=O)CH ), 1.64 (dd, J=12.2, 12.2,
(ESI): m/z calcd for C42
H
79
N
O22SNa [M+Na] : 1032.4774; found:
3
2
3
H, Sia-H3b), 1.41–1.53 (m, 4H, -CH -), 1.10–1.30 ppm (m, 12H,
1032.4840.
2
1
3
-
CH -); C NMR (100 MHz, D O plus one drop CD OD): d=200.5,
2 2 3
1
7
6
3
2
77.2, 175.8, 174.3, 161.2, 104.2, 103.0, 101.2, 80.6, 75.6, 75.6, 74.6,
3.7, 73.5, 73.3, 72.7, 71.8, 70.9, 70.8, 70.7, 70.6, 70.5, 70.4, 70.0,
9.6, 69.5, 69.3, 69.3, 64.4, 63.6, 61.2, 60.3, 52.8, 41.0, 40.4 (ꢂ2),
9.9, 36.8, 31.2, 30.1 (ꢂ2), 30.0 (ꢂ2), 29.9, 29.8, 29.7, 29.4, 26.6,
Methods
À
3.0 ppm; HRMS (ESI): m/z calcd for C H N O S [MÀH] :
49
87
4
26
Preparation of citrate-protected AuNPs (15 nm diameter)
1179.5329; found: 1179.5316. Compound 24 (85 mg, 0.07 mmol)
All glassware used in the following reaction was cleaned with
a freshly prepared Piranha solution (3 parts 18m H SO , 1 part 30%
was dissolved in anhydrous MeOH (5.0 mL) and NaOMe (4.0 mg,
.07 mmol). After stirring at room temperature for 30 min, the solu-
0
2
4
H O ) and then rinsed thoroughly with deionized water prior to
tion was neutralized with 1m aqueous HCl and concentrated in va-
cuo. Purification by size exclusion column chromatography with
deionized H O on Bio-Gel P-2 furnished thiolated sialyllactoside 10
2
2
use (Caution! Wear suitable protective apparel when handling
highly corrosive acids). Spherical citrate-stabilized AuNPs were pre-
pared through the Turkevich procedure. Briefly, 10 mL of 1% tri-
sodium citrate (100 mg, 0.340 mmol) was added directly to a nitro-
gen-deaerated deionized water (200 mL) solution of HAuCl ·3H O
2
[57]
(
72 mg, 0.06 mmol, 88%). R =0.24 (EtOAc/MeOH/H O/AcOH=3/
f 2
2
D
9
1
2
.5/0.5/0.5); [a] =À6.84 (c=0.5 in H O); H NMR (400 MHz, D O):
2
2
d=4.38 (d, J=7.7, 1H, Gal-H1), 4.29 (d, J=7.7, 1H, Glc-H1), 3.88–
.99 (m, 1H, -OCH -), 3.31–3.88 (m, 35H), 3.11–3.30 (m, 7H,
4
2
(44.5 mg, 0.113 mmol) under vigorous stirring at reflux. The solu-
3
2
tion was heated at reflux for an additional 15 min, and the solution
became purple, and finally turned dark red. The resulting solution
was cooled gradually to room temperature, and the UV/Vis absorp-
tion spectrum indicated an SPR band centered at 520Æ1 nm (see
the Supporting Information, FigureS1). The average size of the
metal core of these glyco-AuNPs was 15.2Æ1.5 nm, as determined
by TEM (Figure S1). The citrate-protected AuNPs (pH 5.87) could be
stored at 48C for several months or until use in the ligand-ex-
change reaction with thiolated carbohydrates (see below).
-
NHCH -, Sia-H5), 2.57 (dd, J=4.1, 12.2, 1H, Sia-H3a), 2.40 (t, J=7.0,
H, -CH S-), 2.11 (t, J=7.0, 2H, -(C=O)CH , 1.89 (s, 3H, -(C=O)CH ),
2 2 3
2
2
1
1
.60 (dd, J=12.2, 12.2, 1H, Sia-H3b), 1.38–1.53 (m, 4H, -CH -), 1.06–
2
13
.33 ppm (m, 12H, -CH -); C NMR (100 MHz, D O plus one drop
2
2
CD OD): d=178.1, 175.9, 174.4, 161.3, 104.2, 103.0, 101.3, 80.6,
3
7
7
4
2
5.6, 75.6, 74.7, 73.7, 73.5, 73.3, 72.8, 71.8, 70.8 (ꢂ2), 70.7, 70.6,
0.5, 70.5, 70.4, 69.9, 69.7, 69.5, 69.3 (ꢂ2), 64.5, 63.6, 61.2, 52.8,
1.1, 40.4 (ꢂ2), 39.9, 36.7, 34.1, 29.7 (ꢂ2), 29.4, 29.2, 29.2, 28.6,
6.4, 24.8, 23.0 ppm; HRMS (ESI): m/z calcd for C H N O S
47
85
4
25
À
[MÀH] : 1137.5224; found: 1137.5216.
Synthesis of compound 11
General procedures for the ligand-exchange reaction and the
purification of glyco-AuNPs
Palladium on carbon (Pd/C; 50 mg; 10% Pd content) was suspend-
ed in a solution of 19 (220 mg, 0.33 mmol) in MeOH (15 mL). The
flask was purged with H , and a hydrogen balloon was attached.
Glyco-AuNPs were synthesized in a one-step ligand-exchange reac-
2
After stirring for 12 h at ambient temperature, the solution was fil-
tered through a pad of Celite. The Celite bed was washed with
MeOH, and the combined filtrates were concentrated under re-
duced pressure to provide the amine. Triethylamine (0.138 mL,
tion between citrate-stabilized AuNPs and thiolated glycans (1–4
[25,26]
and 6–11, respectively) following established procedures.
Brief-
ly, reduced thiolated glycans 1–4 (0.004 mmol) were added to a so-
lution of nitrogen-deaerated citrate-protected AuNPs (40 mL). The
pH of the solution was adjusted to 11 with a 0.1m NaOH aqueous
solution. The resulting mixture was stirred at room temperature for
48 h. The glyco-AuNPs were purified by centrifugation at 8000ꢂg,
and the gold sol obtained was redispersed in either NaCl-free
buffer (Tris buffer, 10 mm, pH 7.4) or NaCl-containing phosphate-
buffered saline (PBS, pH 7.4). The centrifugation and redispersion
processes were repeated three times to remove any excess thiolat-
ed glycans or other reactants. UV/Vis absorption spectra were then
acquired to examine NP aggregation to ensure that the NPs did
not aggregate during modification.
0
.99 mmol) and linker 13 (220.0 mg, 0.4 mmol) were added to
a cold solution (À208C) of amine in a mixture of methanol and di-
chloromethane (15.0 mL, 1:1). The resulting mixture was warmed
slowly to room temperature and stirred for 2 h until the starting
material disappeared completely (determined by TLC). Following
removal of volatiles in vacuo, the crude thioacetate (25) was redis-
solved in 1m aqueous NaOH solution (20.0 mL). After stirring at
room temperature for 30 min, the solution was neutralized with
1
m aqueous HCl and concentrated in vacuo. Purification by size
k
exclusion chromatography on BioGel P-2 gel furnished thiolated P
Chem. Eur. J. 2015, 21, 3956 – 3967
www.chemeurj.org
3965
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Preparation of mixed monolayer-protected glyco-AuNPs (5–
AuNP to 11–AuNP)
Fabrication of the oriented immobilized antibody slide
Nexterion H glass slides were treated with a 50 mm solution of 26
in PBS (pH 8.5) at room temperature for 24 h to derivatize the N-
hydroxysuccinimide (NHS)-ester groups. The slides were washed
with deionized water and then dried by centrifugation. The un-
reacted NHS-esters were deactivated by immersing the slide into
a solution containing ethanolamine (25 mm) in borate buffer
(100 mm, pH 9.2) at room temperature for 1 h. The slides were
washed sequentially with phosphate-buffered saline (PBS, pH 7.4)
and PBS containing 0.05% Tween 20 (PBST), rinsed with deionized
water, and dried prior to use. Monoclonal capture antibodies
(6.6 mm in PBS buffer, pH 7.4) were dispensed by using a robotic
contact arrayer (AD1500 Arrayer, BioDot) fitted with Stealth Pins
SMP3 (ArrayIt Corporation) onto 26-functionalized glass slides. The
printing process was performed at a relative humidity of 94%, and
the temperature was maintained below 268C. The slide was al-
lowed to react at 48C for 24 h; this was followed by rinsing with
For the preparation of 5–AuNP to 11–AuNP, a mixed thiol solution
was prepared from the corresponding thiolated glycans (70 mol%)
and linker 12 (30 mol%). The ligand-exchange reaction of the
AuNPs with mixed thiols followed the general functionalization
method described above, with the exception that the pure thiol
was replaced with the mixed thiol solution for the treatment of
the citrate-stabilized AuNPs. The remainder of the coupling was
performed following the same method described in detail above.
The concentration of the glyco-AuNP solution, approximately
2
0 nm, was determined by measuring the SPR absorption and
[58]
8
À1
À1
using a reported
extinction coefficient (3.64ꢂ10 m cm ) to
provide 5–AuNP, 6–AuNP, 7–AuNP, 8–AuNP, 9–AuNP, 10–AuNP, and
1–AuNP. These glyco-AuNPs could be stored at 48C for several
1
months until use.
deionized H O to remove any unbound antibodies. The remaining
2
BAs on the surface were then blocked with dextran (100 mm) con-
taining 1% BSA in water at room temperature for 2 h to prevent
nonspecific adhesion/interactions. Following a wash, the oriented-
antibody immobilized microarray slides could be probed with cog-
nate lectin and glyco-AuNPs either in a layer-by-layer format or
in situ agglutination for detection with the naked eye (see below).
Stability of glyco-AuNPs in NaCl-containing aqueous buffer
The effect of NaCl on the stability of AuNPs was studied by a colori-
[26]
metric assay following a protocol described previously. Thiol-ter-
minated lactose derivatives 1–4 (0.1 mm) were allowed to react
with citrated-protected AuNPs (5 nm) for 48 h. The resulting lac-
tose-functionalized AuNPs were divided (200 mL) into 96-well mi-
crotiter plates without purification. NaCl was added to the wells to
obtain final concentrations of 0, 12.5, 25, 50, 100, 150, 300, and
In situ lectin-induced AuNP agglutination on microarrays and
naked-eye detection through silver enhancement
6
00 mm. Following a 1 h incubation, the absorbance at 620 nm
e
was recorded with an ELISA plate reader (SpectraMax M2 , Molecu-
lar Devices).
The oriented immobilized antibody slide was incubated with
1
00 mL of a solution containing a mixture of lectin and glyco-
AuNPs (5 nm) for 2–8 h at room temperature. A 16-well incubation
chamber (Whatman, Inc.) was used to divide the slide into 16
blocks (each with a 9ꢂ9 mm format) to prevent cross-contamina-
tion when different samples were applied to neighboring blocks.
After incubation, the unbound materials were removed by sequen-
Nonspecific protein adsorption on glyco-AuNPs
BSA was used to measure nonspecific protein adsorption on glyco-
AuNPs. Glyconanoparticles (such as 5–AuNP, 3 nm, 1 mL) were in-
cubated with varying concentrations of BSA (0, 0.02, 2, 200 mm) at
room temperature for 8 h. The solution was monitored by UV/Vis
spectrophotometry and a particle size analyzer.
tial washing with PBST, PBS, and deionized H O for 5 min each. Fi-
2
nally, the glass slide was dried by centrifugation, and the bound
AuNPs were further visualized by silver enhancement for 14 min
unless otherwise noted. In silver enhancement, the AuNPs act as
the nucleation sites for metallic silver, enabling visual detection of
the darkened, enlarged sites.
Plate-based colorimetric detection of lectins by glyco-AuNPs
In regular 96-well microtiter plates, solutions of glyco-AuNPs (3 nm)
in PBS buffer containing 0.005% Tween 20 (PBST, 200 mL) were in- Acknowledgements
cubated with cognate lectins (500 nm) at room temperature for
2
7
h. Following incubation, the change in absorbance from 450–
50 nm was measured with an ELISA plate reader as described
This work was supported financially by the National Tsing Hua
University, Academia Sinica, and the Ministry of Science and
Technology of Taiwan.
[26]
above.
Fabrication of the random-immobilized antibody slide
Keywords: antibodies
nanoparticles · lectins
·
biosensors
·
boronic acid
·
For the immobilization of antibodies on a glass slide, microarrays
were generated using a AD1500 arrayer (contact mode) to print
antibodies (6.7 mm in PBS buffer, pH 7.4) onto a NEXTERION Slide H
with a contact time of 300 ms. The resulting slides were incubated
at 48C overnight. Then, all of the remaining unreacted activated
ester groups were capped by incubation with a capping reagent
[1] C. Fasting, C. A. Schalley, M. Weber, O. Seitz, S. Hecht, B. Koksch, J. Der-
nedde, C. Graf, E. W. Knapp, R. Haag, Angew. Chem. Int. Ed. 2012, 51,
1
0472–10498; Angew. Chem. 2012, 124, 10622–10650.
2] Lectins, 2nd ed. (Eds.: N. Sharon, H. Lis), Kluwer Academic, Dordrecht,
003.
[
2
(
see general information) at 258C for 1 h. After incubation, the
[
[
3] W. I. Weis, K. Drickamer, Annu. Rev. Biochem. 1996, 65, 441–473.
4] J. Stevens, O. Blixt, L. Glaser, J. K. Taubenberger, P. Palese, J. C. Paulson,
I. A. Wilson, J. Mol. Biol. 2006, 355, 1143–1155.
glass slide was washed sequentially with PBST, PBS, and deionized
H O for 5 min each and finally dried by centrifugation. The anti-
2
body slide was stored in the proper chamber at 48C prior to fur-
ther use.
[5] S. Fukuta, J. L. Magnani, E. M. Twiddy, R. K. Holmes, V. Ginsburg, Infect.
Immun. 1988, 56, 1748–1753.
Chem. Eur. J. 2015, 21, 3956 – 3967
www.chemeurj.org
3966
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
6] T. B. Geijtenbeek, D. S. Kwon, R. Torensma, S. J. van Vliet, G. C. van Duijn-
hoven, J. Middel, I. L. Cornelissen, H. S. Nottet, V. N. KewalRamani, D. R.
Littman, C. G. Figdor, Y. van Kooyk, Cell 2000, 100, 587–597.
[35] S. K. Podder, A. Surolia, B. K. Bachhawat, Eur. J. Biochem. 1974, 44, 151–
160.
[36] M. Huet, Eur. J. Biochem. 1975, 59, 627–633.
[
[
[
7] L. L. Kiessling, R. A. Splain, Annu. Rev. Biochem. 2010, 79, 619–653.
8] J. E. Turnbull, R. A. Field, Nat. Chem. Biol. 2007, 3, 74–77.
9] D. A. Giljohann, C. A. Mirkin, Nature 2009, 462, 461–464.
[37] G. Bains, R. T. Lee, Y. C. Lee, E. Freire, Biochemistry 1992, 31, 12624–
12628.
[38] K. Matsumura, K. Higashida, H. Ishida, Y. Hata, K. Yamamoto, M. Shigeta,
Y. M. Horikawa, X. Wang, E. Miyoshi, J. Gu, N. Taniguchi, J. Biol. Chem.
2007, 282, 15700–15708.
[39] W.-C. Wang, R. D. Cummings, J. Biol. Chem. 1988, 263, 4576–4585.
[40] N. Shibuya, I. J. Goldstein, W. F. Broekaert, M. Lubaki, B. Peters, J. Biol.
Chem. 1987, 262, 1596–1601.
[41] N. Gilboa-Garber, D. Sudakevitz, M. Sheffi, R. Sela, C. Levene, Glycoconju-
gate J. 1994, 11, 414–417.
[
[
[
10] K. Saha, S. S. Agasti, C. Kim, X. Li, V. M. Rotello, Chem. Rev. 2012, 112,
739–2779.
11] M. Marradi, F. Chiodo, I. Garcia, S. Penadꢁs, Chem. Soc. Rev. 2013, 42,
728–4745.
12] J. M. de la Fuente, A. G. Barrientos, T. C. Rojas, J. Rojo, J. Canada, A. Fer-
nandez, S. Penadꢁs, Angew. Chem. Int. Ed. 2001, 40, 2257–2261; Angew.
Chem. 2001, 113, 2317–2321.
2
4
[
[
[
[
[
13] C.-C. Lin, Y.-C. Yeh, C.-Y. Yang, C.-L. Chen, G.-F. Chen, C.-C. Chen, Y.-C.
Wu, J. Am. Chem. Soc. 2002, 124, 3508–3509.
14] A. G. Barrientos, J. M. de La Fuente, T. C. Rojas, A. Fernandez, S. Penadꢁs,
Chem. Eur. J. 2003, 9, 1909–1921.
[42] X. Song, H. Yu, X. Chen, Y. Lasanajak, M. M. Tappert, G. M. Air, V. K.
Tiwari, H. Cao, H. A. Chokhawala, H. Zheng, R. D. Cumming, D. F. Smith,
J. Biol. Chem. 2011, 286, 31610–31622.
[43] J. C. Manimala, T. A. Roach, Z. T. Li, J. C. Gildersleeve, Angew. Chem. Int.
Ed. 2006, 45, 3607–3610; Angew. Chem. 2006, 118, 3689–3692.
[44] B. P. Peters, S. Ebisu, I. J. Goldstein, M. Flashner, Biochemistry 1979, 18,
5505–5511.
[45] E. Song, J. Li, H. Wei, Y. Song, Anal. Methods 2012, 4, 1199–1201.
[46] M.-L. Chen, A. K. Adak, N.-C. Yeh, W.-B. Yang, Y.-J. Chuang, C.-H. Wong,
K.-C. Hwang, J.-R. Hwu, S.-L. Hsieh, C.-C. Lin, Angew. Chem. Int. Ed. 2008,
47, 8627–8630; Angew. Chem. 2008, 120, 8755–8758.
[47] P.-C. Lin, S.-H. Chen, M.-L. Chen, A. K. Adak, Y.-J. Chen, C.-C. Lin, Anal.
Chem. 2009, 81, 8774–8782.
15] C. L. Schofield, A. H. Haines, R. A. Field, D. A. Russell, Langmuir 2006, 22,
6707–6711.
16] S. Takae, Y. Akiyama, H. Otsuka, T. Nakamura, Y. Nagasaki, K. Kataoka,
Biomacromolecules 2005, 6, 818–824.
17] C. L. Schofield, B. Mukhopadhyay, S. M. Hardy, M. McDonnell, R. A. Field,
D. A. Russell, Analyst 2008, 133, 626–634.
[
18] A. K. Adak, H.-J. Lin, C.-C. Lin, Org. Biomol. Chem. 2014, 12, 5563–5573.
19] C.-C. Lin, Y.-C. Yeh, C.-Y. Yang, G.-F. Chen, Y.-C. Che, Y.-C. Wu, C.-C. Chen,
Chem. Commun. 2002, 2920–2921.
[
[
[
[
20] Y. Zhang, S. Luo, Y. Tang, L. Yu, K.-Y. Hou, J.-P. Cheng, X. Zheng, P. G.
[48] A. K. Adak, B.-Y. Li, L.-D. Huang, T.-W. Lin, T.-C. Chang, K.-C. Hwang, C.-C.
Lin, ACS Appl. Mater. Interfaces 2014, 6, 10452–10460.
[49] P.-C. Lin, P.-H. Chou, S.-H. Chen, H.-K. Liao, K.-Y. Wang, Y.-J. Chen, C.-C.
Lin, Small 2006, 2, 485–489.
[50] J.-A. Ho, W.-L. Hsu, W.-C. Liao, J.-K. Chiu, M.-L. Chen, H.-C. Chang, C.-C.
Lin, Biosens. Bioelectron. 2010, 26, 1021–1027.
Wang, Anal. Chem. 2006, 78, 2001–2008.
21] J. M. de La Fuente, P. Eaton, A. G. Barrientos, M. Menendez, S. Penadꢁs,
J. Am. Chem. Soc. 2005, 127, 6192–6197.
22] K. El-Boubbou, D. C. Zhu, C. Vasileiou, B. Borhan, D. Prosperi, W. Li, X.
Huang, J. Am. Chem. Soc. 2010, 132, 4490–4499.
[
[
[
23] X. Wang, O. Ramstrçm, M. Yan, Anal. Chem. 2010, 82, 9082–9089.
24] X. Wang, O. Ramstrçm, M. Yan, Analyst 2011, 136, 4174–4178.
25] Y.-Y. Chien, M.-D. Jan, A. K. Adak, H.-C. Tzeng, Y.-P. Lin, Y.-J. Chen, K.-T.
Wang, C.-T. Chen, C.-C. Lin, ChemBioChem 2008, 9, 1100–1109.
26] C.-C. Yu, L.-D. Huang, D. H. Kwan, W. W. Wakarchuk, S. G. Withers, C.-C.
Lin, Chem. Commun. 2013, 49, 10166–10168.
[51] V. Wittmann, R. J. Pieters, Chem. Soc. Rev. 2013, 42, 4492–4502.
[52] H. Kato, H. Uzawa, T. Nagatsuka, S. Kondo, K. Sato, I. Ohsawa, M. Kana-
mori-Kataoka, Y. Takei, S. Ota, M. Furuno, H. Dohi, Y. Nishida, Y. Seto, Car-
bohydr. Res. 2011, 346, 1820–1826.
[53] L. Deng, O. Norberg, S. Uppalapati, M. Yan, O. Ramstrom, Org. Biomol.
Chem. 2011, 9, 3188–3198.
[54] F. Sallas, S.-I. Nishimura, J. Chem. Soc. Perkin Trans. 1 2000, 2091–2103.
[55] F. Morvan, A. Meyer, A. Jochum, C. Sabin, Y. Chevolot, A. Imberty, J.-P.
Praly, J.-J. Vasseur, E. Souteyrand, S. Vidal, Bioconjugate Chem. 2007, 18,
1637–1643.
[56] S.-P. Li, W.-C. Hsiao, C.-C. Yu, W.-T. Chien, H.-J. Lin, L.-D. Huang, C.-H. Lin,
W.-L. Wu, S.-H. Wu, C.-C. Lin, Adv. Synth. Catal. 2014, 356, 3199–3213.
[57] B. V. Enꢃstꢃn, J. Turkevich, J. Am. Chem. Soc. 1963, 85, 3317–3324.
[58] W. Lian, D. Wu, D. V. Lim, S. Jin, Anal. Biochem. 2010, 401, 271–279.
[
[
27] A. J. Reynolds, A. H. Haines, D. A. Russell, Langmuir 2006, 22, 1156–
1163.
[
[
28] E. Mabon, T. Astrup, M. Barboiu, Chem. Commun. 2010, 46, 5491–5493.
29] E. Ostuni, R. G. Chapman, M. N. Liang, G. Pier, D. E. Ingber, G. M. White-
sides, Langmuir 2001, 17, 6336–6343.
[
[
30] T.-H. Yang, L.-D. Huang, Y.-W. Harm, C.-C. Lin, J.-K. Chang, C.-I. Wu, J.-M.
Wu, Small 2013, 9, 3169–3182.
31] J. J. Storhoff, A. A. Lazarides, R. C. Mucic, C. A. Mirkin, R. L. Letsinger,
G. C. Schatz, J. Am. Chem. Soc. 2000, 122, 4640–4650.
[
[
32] D. C. Hone, A. H. Haines, D. A. Russell, Langmuir 2003, 19, 7141–7144.
33] D. J. Vanderah, J. Arsenault, H. La, R. S. Gates, V. Silin, C. W. Meuse, Lang-
muir 2003, 19, 3752–3756.
[34] M. Zheng, F. Davidson, X. Huang, J. Am. Chem. Soc. 2003, 125, 7790–
Received: October 21, 2014
7791.
Published online on January 8, 2015
Chem. Eur. J. 2015, 21, 3956 – 3967
www.chemeurj.org
3967
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim